Guidelines and recommendations to better safeguard nuclear reactors from extreme natural events
Probabilistic safety assessment (PSA) is a method that quantifies risk measures in NPPs. It determines what undesired scenarios can occur, their likelihood and potential outcomes. If the results of a PSA conclude that low-probability events can lead to extreme consequences, system suppliers, utilities or safety authorities can take appropriate action to reinforce the defence-in-depth of the plant. This concept contains several levels of protection to safeguard public health and safety from hazards associated with nuclear materials. The EU-funded ASAMPSA_E(opens in new window) (Advanced safety assessment: Extended PSA) project promoted good practices to identify critical situations with the help of PSAs and to define appropriate criteria for decision-making in Europe. To achieve its goals, ASAMPSA_E focused on extended PSA, which applies to one or more NPP sites and their environments. Extended PSA calculates the risk generated by main radioactivity sources, including reactor core and spent fuel storages, on the site. It considers all operating states for each main source, and all possible relevant accident-initiating internal and external events that affect one NPP or the whole site. Project partners began by identifying PSA end-user needs for extended PSA. They developed a set of recommendations, guidance reports and guidelines that were externally reviewed. In total, 24 technical reports were submitted by over 100 experts from 31 organisations in 18 European countries, Japan and the United States. These mainly include general PSA issues such as Fukushima nuclear accident lessons learnt for PSA, external hazards to be considered, and a methodology for selecting initiating events and hazards for consideration in an extended PSA. Also included are guidelines for practices to model and implement various external hazards in extended PSA, and PSA and severe accident management strategies. At a technical level, ASAMPSA_E proposed a new framework to discuss how extended PSA can be developed efficiently and implemented successfully in order to verify if the robustness of NPPs in their environments is sufficient.
